Utilizing YoLOv8 and LSTM for Worker Mobility Tracker-Based Helmet and Equipment Detection in the Mining Industry

Abstract

Ensuring staff safety in mining operations requires close monitoring of PPE and compliance with safety regulations.
However, due to the unpredictable and hazardous nature of mining circumstances, it is not feasible to send people
below. Two essential components of worker safety—real-time detection of helmets and PPE as well as worker
mobility—remains a challenge for most current systems. This project's strategy for overcoming these issues is to
improve security and monitoring via the use of machine learning technology. A state-of-the-art object identification
model called YOLOv8 is used by the system to allow real-time detection of workers' helmets and personal
protective equipment (PPE). In addition, the work makes use of LSTM, a subset of RNN, to monitor employee
actions both in the present and in the past. Whether it's dangerous motions or ones caused by exhaustion, the system
can identify them at 10-millisecond intervals and take preventative actions to ensure the user's safety. By integrating
YOLOv8 and LSTM, the proposed approach eliminates room for human mistake and boosts workplace safety to an
impressive 95.9%. This cutting-edge technology is revolutionizing mining safety solutions by enhancing operating
efficiency and offering continuous monitoring. The following terms are used as keywords: helmet detection, PPE,
worker mobility tracking, RNN, LSTM, YOLOv8.